Julia Tutorial - Why Should I Learn Julia Programming Language

Key Takeaways

Hello and welcome to core basics coding tutorial. Today we are going to talk about this new programming language called Julia. It's a new language that's making a lot of buzz nowadays and it is invented by some of the brilliant people at MIT. So the first question that comes to anyone's mind is there are so many languages out there in the world. Why do I need to learn yet another programming language? Well, that's a good question and one line answer to that question is Julia combines Python's convenience with C's performance. And I don't mean Python and C literally. What I really mean is it combines convenience of dynamic languages with the performance of statically linked languages. Now what people have been doing is whenever they need execution speed or performance they will use C because it C code runs crazy fast. But the problem is the code development time is significantly higher. I have done software development for last 15 years using C and Python and I have always found that C code development is very very slow. Compared to that if you are writing code in Python you are minimum five times faster. So it gives it Python has an age there whenever whenever it comes to productivity but the problem is the it suffers from the performance. Now don't get me wrong it works reasonably well on most of the projects but whenever you want a super fast performance it sometimes fails. In this kind of situation, what people do is they write their code initially. They build their building blocks uh using a dynamic language such as Python. And using dynamic language, you can always do uh fast prototyping. Now once you have your code running, you identify the slow running parts of your code and you move them into compile language such as C and Forrun. If you're still concerned about performance, you move the slow running part of your compile language code into an assembly language. Now this works great but the problem here is the transition between these layers is extremely tough. The barrier is high. You have to write C extension if you are writing a code in Python and that takes considerable amount of time. Not only that, uh you need a programmer who knows both Python and C in order to work on this project. So this is a big problem. It consumes so much time and it will increase your budget on your project. Julia breaks this resistance. Here is the sample code. The first one is a function written in Julia and you can run code native call on it to examine how the native code looks like and based on that you can optimize the performance. Now all of these you can do within Julia in a single environment and that's pretty awesome. Julia comes with all these features. I have listed only few. It has many more features into it but just to quickly go over these features. Number one uh selling point for Julia is the performance similar to C. It has a very sophisticated dynamic type system as well. So your productivity is still high as high as Python. It uses git compilation using LLVM compiler framework and that's one of the reason why it is very fast. Other than that, it supports multiple dispatch and it is designed for parallelism and cloud computing. Here are some of the benchmarking results uh taken from uh jiuliang.org website. They ran these bank benchmarks on some of the basic functions. So just look at this quick shot. So if you're running quickshot uh here the number one means the performance of C. So Julia gives 1.15 which is very similar to C. If you compare that with N mat lab it is super slow 264. Python is also slow and JavaScript and Java are reasonably faster but still Julia clearly wins here. You can go to Julia Lang Lang Julia.org website and look at the complete chart for all the benchmarking results. Here is the quote that I took from Ivan Miller's website. I like this very much. It says it's poised to do for technical computing what Node.js is doing for web development. getting disperate groups of programmers to code in the same language. With the node.js, it was the front-end designers and the backend developers. With Julia, it is the domain experts and the speed fricks. That is a major accomplishment. Julia lang.org is an official website. So if you are uh the technical computing enthusiast who is interested in fast prototyping as well as performance then Julia is definitely a language that you should check out. Thanks.